A flock-tailored early warning system for low pathogenic avian influenza (LPAI) in commercial egg laying flocks

The aim of this study was to develop and evaluate an early warning system (EWS) for commercial egg laying flocks to detect the subtle mortality and egg production changes that characterize low pathogenic avian influenza virus (LPAIV) infections. An EWS will create an alert when the recommended 'trigger point' is reached or exceeded. Previously used EWSs are based on fixed alert levels, while the proposed EWS customizes the alert level to each flock. While a fixed approach may be valid for highly pathogenic diseases, it results in a lower detection probability for low pathogenic diseases. The EWS was based on daily data collected from flocks affected by the 2000-2004 H6N2 LPAI epidemic in California. Three EWSs were evaluated: (1) EWS1, which is triggered when the observed mortality increase or production decrease exceeds more than 'x' times the expected daily value (2.75-3.50 times the expected mortality), (2) EWS2, which is triggered when the observed mortality increase or production decrease exceeds more than "y" times during each of 2 consecutive days the expected daily values (1.75-2.15 times the expected mortality), and (3) a combination of the two. The EWSs were evaluated according to three parameters: detection delay (days) of a LPAI outbreak, false alerts (%) and outbreaks missed (%). Results showed that an egg production-based EWS added no benefit to a mortality-based system, mainly because H6N2 LPAI-related egg production decrease always occurred after increase in mortality. Combining the two EWSs resulted in a reduced detection delay and no missed outbreaks, but at the expense of a slight increase in the number of false alerts triggered. The system presented in this study also outperformed fixed EWSs in all three evaluated parameters. The proposed EWS, if used as part of a poultry cooperative program and combined with a rapid laboratory diagnosis, could be a useful tool in the detection and control of LPAI outbreaks and other poultry diseases. Built in a spreadsheet, the system could be inexpensively, easily and quickly incorporated into a commercial egg production farm decision support system. In addition, the proposed system could be quickly adjusted to changing epidemic situations, and easily customized to individual flocks. (C) 2009 Elsevier B.V. All rights reserved.